There are words that should be considered easy to understand in conversation, that get a blank-stare response. Recently, a friend of mine gave me the example of the word "antiquated". I was surprised by the fact that this word was questioned and denounced as "too big a word" (a label which is often used for a word that someone doesn't understand). There seems to be a lack of contextual inference. For example, the sentence "Young people have trouble understanding older generations due to the antiquated nature of their language" gives us very good hints as to the meaning of antiquated.
Secondly, there seems to be a lack of meaning-deduction from the word structure. Many words, like antiquated, contain roots that are easily recognizable. If a sentence without contextual information was used, there still is a possibility of inferring the meaning of the word by it's structure. For example, in a conversation where one is describing a record player, the listener may respond "that's antiquated." The root antiquate- and it's similarity to antique may guide the first converser too infer that the listener has just said "that's old-fashioned" or "I really don't care what your saying because what you've just described is something I threw out ten years ago" (the two meanings, of course, depend on the tone used when the response is uttered).
There are other types of words that have come up in conversation that have been questioned. These words have unfamiliar roots. I used the word "paradox" a while ago. Now the receiving converser had never heard this word before. There could be a few reasons for not knowing this words. First, she was a college freshman, a new one at that. Second, and this only occurs to me know, is that a paradox is more often than not used as scientific and mathematic lingo. It is still a very descriptive word for the types of situation it describes.
(A little context around how the word paradox came up in said conversation. It was a question of why there were less men then women in the ballroom class. My response to this was that dancing is a bit of a paradox with guys. They don't want to embarrass themselves by looking like a dancing monkey, and therefore drive away the women on the dance floor. On the other hand, they know that women like guys who can dance, and therefore the should dance to attract women on the dance floor. This is a paradox in that they attempt to do neither. So the resort to the common method of restoring balance and sanity to a paradoxical universe: they drink large sums of alcohol).
This decline of language can be attributed to two causes. First is newsprint (and via evolution, webprint). The language used is meant for the least common denominator which is the fifth-grade reading level. Apparently, in these days of much-lauded need for higher education journalism does not assume that anyone has made it out of the fifth grade. I think print media has had these accusations thrown it's way ever since the first English-speaking newspaper rolled out in 1622,
The second reason, and more powerfully so, is television.
The Department of Education states that language skills are best developed through reading and interactions with others in conversation and play. Excessive television watching can impede this development. Hours spent watching TV make risk-taking and social relationships difficult for many children. Television viewing is up, much of which (and here I start to make some rash generalizations, unfounded by the fact that I am neither a sociologist, nor a parent) is probably due to the rise of the two-income household. In many of these households, the time spent between when the child gets home and the first of the parents get home is spent watching TV or playing video games. There is little physical activity as well as very little social interaction involved while TV watching is in progress. The social interaction, especially at a young age, is very important for developing good conversation and language skills. Much of a child's early language development is learned through example.
When it comes to conversation, though TV is not the only detractor. I've encountered one child in particular whose language skills were very weak, in terms of pronunciation. Her mother was a stay-at-home mom, who was also very careful about regulating the child's TV intake. However (and this may be due to a certain amount of shyness), her pronunciation of words was terrible since she lacked interaction with other children. Her pronunciation was never corrected by her parents because they understood her; there was no need to correct it. Essentially she had developed her own dialect. Needless to say, I felt like I was talking to someone from rural New Zealand, or a Cockney Londoner: I required a translator.
I find that fact that I'm interested in the use of the English language very amusing. I'm a mathematician and computer scientist by trade. Almost as a rule, we have very little formal training in written (and verbal) communications. Many of us do a lot of reading, which is a great way to research writing and language usage. The only thing missing from this sort of research are the formal rules that back things up (like the use of it's versus its, something I needed explained to me to start using them correctly). Still, we're also very good emulators. Much of our sciences are learn-by-example types of sciences. Another reason that we're a lot like little children.
 The first English language news pamphlet was The Weekly Newes, published in London in 1622. This wasn't a daily, but was printed whenever there was an interesting event. The first daily was the London Gazette, published in 1666, which is widely considered the first true newspaper. A Brief History of Newspapers, Phil Barber
 How Television Viewing Affects Children, University of Maine Cooperative Extension